arXiv Quantitative Biology @arxiv_bio@qoto.org

Deciphering the neural mechanisms that transform sensory experiences into meaningful semantic representations is a fundamental challenge in cognitive neuroscience. While neuroimaging has mapped a distributed semantic network, the format and neural code of semantic content remain elusive, particularly for complex, naturalistic stimuli. Traditional brain decoding, focused on visual reconstruction, primarily captures low-level perceptual features, missing the deeper semantic essence guiding human cognition. Here, we introduce a paradigm shift by directly decoding fMRI signals into textual descriptions of viewed natural images. Our novel deep learning model, trained without visual input, achieves state-of-the-art semantic decoding performance, generating meaningful captions that capture the core semantic content of complex scenes. Neuroanatomical analysis reveals the critical role of higher-level visual regions, including MT+, ventral stream visual cortex, and inferior parietal cortex, in this semantic transformation. Category-specific decoding further demonstrates nuanced neural representations for semantic dimensions like animacy and motion. This text-based decoding approach provides a more direct and interpretable window into the brain's semantic encoding than visual reconstruction, offering a powerful new methodology for probing the neural basis of complex semantic processing, refining our understanding of the distributed semantic network, and potentially inspiring brain-inspired language models.

Many rare genetic diseases exhibit recognizable facial phenotypes, which are often used as diagnostic clues. However, current facial phenotype diagnostic models, which are trained on image datasets, have high accuracy but often suffer from an inability to explain their predictions, which reduces physicians' confidence in the model output.In this paper, we constructed a dataset, called FGDD, which was collected from 509 publications and contains 1147 data records, in which each data record represents a patient group and contains patient information, variation information, and facial phenotype information. To verify the availability of the dataset, we evaluated the performance of commonly used classification algorithms on the dataset and analyzed the explainability from global and local perspectives. FGDD aims to support the training of disease diagnostic models, provide explainable results, and increase physicians' confidence with solid evidence. It also allows us to explore the complex relationship between genes, diseases, and facial phenotypes, to gain a deeper understanding of the pathogenesis and clinical manifestations of rare genetic diseases.

Species complexes are groups of closely related populations exchanging genes through dispersal. We study the dynamics of the structure of species complexes in a class of metapopulation models where demes can exchange genetic material through migration and diverge through the accumulation of new mutations. Importantly, we model the ecological feedback of differentiation on gene flow by assuming that the success of migrations decreases with genetic distance, through a specific function $h$. We investigate the effects of metapopulation size on the coherence of species structures, depending on some mathematical characteristics of the feedback function $h$. Our results suggest that with larger metapopulation sizes, species form increasingly coherent, transitive, and uniform entities. We conclude that the initiation of speciation events in large species requires the existence of idiosyncratic geographic or selective restrictions on gene flow.

Setting up a psychedelic study can be a long, arduous, and kafkaesque process. This rapidly-developing field poses several unique challenges for researchers, necessitating a range of considerations that have not yet been standardised. Many of the complexities inherent to psychedelic research also challenge existing assumptions around, for example, approaches to psychiatric prescribing, the conceptual framing of the placebo effect, and definitions of selfhood. This review paper brings together several of the major psychedelic research teams across the United Kingdom to formalise these unique considerations, identify continuing areas of debate, and provide a practical, experience-based guide, with recommendations for policymakers and future researchers intending to set up a psychedelic research study or clinical trial. We approach this such that the paper can either be read end to end, or treated as a manual: readers can dip into relevant sections as needed.

Fermentation of organisms for recycling is important for the efficient cycling of nitrogen and phosphorus resources for a sustainable society, but the functionality of fermented products needs to be evaluated. Here, we clarify the anti-pathogenic properties for fish of a compost-type feed additive fermented by thermophilic Bacillaceae using non-edible marine resources as raw materials. After prior administration of the compost extract to seabream as a fish model for 70 days, the mortality rate after 28 days of exposure to the fish pathogen Edwardsiella reached a maximum of 20%, although the rate was 60% without prior administration. Under such conditions, the serum complement activity of seabream increased, and the recovery time after anesthesia treatment was also fasten. Furthermore, texture and HSV analysis using field photos statistically visualized differences in the degree of smoothness and gloss of the fish body surface depending on the administration. These results suggest that thermophile-fermented compost is effective as a functional feed additive against fish disease infection, and that such soundness can be estimated by body surface analysis. This study provides a new perspective for the natural symbiosis industry, as well as for the utilization of field non-invasive diagnosis to efficiently estimate the quality of its production activities.

Because organisms are able to sense its passage, it is perhaps tempting to treat time as a sensory modality, akin to vision or audition. Indeed, certain features of sensory estimation, such as Weber's law, apply to timing and sensation alike (Gibbon, 1977; Pardo-Vazquez et al., 2019). However, from an organismal perspective, time is a derived feature of other signals, not a stimulus that can be readily transduced by sensory receptors. Its importance for biology lies in the fact that the physical world comprises a complex dynamical system. The multiscale spatiotemporal structure of sensory and internally generated signals within an organism is the informational fabric underlying its ability to control behavior. Viewed this way, temporal computations assume a more fundamental role than is implied by treating time as just another element of the experienced world (Paton & Buonomano, 2018). Thus, in this review we focus on temporal processing as a means of approaching the more general problem of how the nervous system produces adaptive behavior.

The classic heuristic formula for calculating the mean arterial blood pressure (MAP) in the human cardiac cycle using 1/3 of the systolic pressure (P_S) plus 2/3 of the diastolic pressure (P_D) importantly neglects the nonlinear effect of heart rate (HR) on blood pressure. Regression analysis on blood pressure data from patients experiencing a variety of heart rates captured a logarithmic relationship between mean arterial pressure and heart rate, F_S=0.086 *ln(HR-41.2)+0.079, where F_S is the MAP form factor corresponding to the fraction of time the cardiac cycle spends in ventricular systole. The logarithmic relationship is qualitatively distinct from the linear and exponential relationships previously suggested in the literature yet better aligns with the known human physiology. This heart rate-dependent formula for the mean arterial pressure recapitulates the classic intuition as a special case (HR=60bpm implies F_S =1/3). In addition, the model suggests several other MAP heuristics: HR=45bpm implies F_S = 1/5, HR=50bpm implies F_S = 1/4, HR=175bpm implies F_S = 1/2. A calculator for the MAP can be found at https://mngu265.shinyapps.io/map\_calculator/

The symbolic architecture of non-ordinary consciousness remains largely unmapped in cognitive science and artificial intelligence. While conventional models prioritize rational coherence, altered states such as those induced by psychedelics reveal distinct symbolic regimes characterized by recursive metaphor, ego dissolution, and semantic destabilization. We present \textit{Glyph}, a generative symbolic interface designed to simulate psilocybin-like symbolic cognition in large language models. Rather than modeling perception or mood, Glyph enacts symbolic transformation through recursive reentry, metaphoric modulation, and entropy-scaled destabilization -- a triadic operator formalized within a tensorial linguistic framework. Experimental comparison with baseline GPT-4o reveals that Glyph consistently generates high-entropy, metaphor-saturated, and ego-dissolving language across diverse symbolic prompt categories. These results validate the emergence of non-ordinary cognitive patterns and support a new paradigm for simulating altered consciousness through language. Glyph opens novel pathways for modeling symbolic cognition, exploring metaphor theory, and encoding knowledge in recursively altered semantic spaces.

In this study, we present a stochastic simulation model designed to explicitly incorporate cell cycle length, overcoming limitations associated with classical compartmental models. Our approach employs a delay mechanism to represent the cell cycle, allowing the use of arbitrary distributions for cell cycle lengths. We demonstrate the feasibility of our model by fitting it to experimental data from melanoma cell lines previously studied by Vittadello et al. Notably, our model successfully replicates experimentally observed synchronization phenomena that multi-stage models could not adequately explain. By using a gamma distribution to model cell cycle lengths, we achieved excellent agreement between our simulations and empirical data, while significantly reducing computational complexity and parameter estimation challenges inherent in multi-stage approaches. Our results highlight the importance of explicitly incorporating cell cycle lengths in modeling cell populations, with potential implications for optimizing cell cycle-dependent tumor therapies.

Ferroptosis is a form of cell death discovered in recent years, induced by excessive peroxidation of phospholipids. Glutathione peroxidase 4 (GPx4) is an intracellular enzyme that can repair the peroxidized phospholipids on membranes, thus regulating ferroptosis. By combining multiscale molecular dynamics (MD) simulations and experimental assays, we investigate the binding mechanisms of GPx4 on membranes. Using coarse-grained MD simulations, we found that L130 and its adjacent residues on GPx4 can form a stable and unique binding interface with PE/PS-rich and peroxidized membranes. Subsequent all-atom MD simulations verified the stability of the binding interface. The critical residue on the interface, L130, was inserted deeply into the membrane as a hydrophobic anchor and guided the reaction center toward the membrane surface. Enzyme activity assays and in vitro cell experiments showed that mutations of L130 resulted in weaker activities of the enzyme, probably caused by non-functional binding modes of GPx4 on membranes, as revealed by in silico simulations. This study highlights the crucial role of the hydrophobic residue, L130, in the proper anchoring of GPx4 on membranes, the first step of its membrane-repairing function.

Horizontal gene transfer (HGT) is an important process in bacterial evolution. Current phylogeny-based approaches to capture it cannot however appropriately account for the fact that HGT can occur between bacteria living in different ecological niches. Due to the fact that arboreal networks are a type of multiple-rooted phylogenetic network that can be thought of as a forest of rooted phylogenetic trees along with a set of additional arcs each joining two different trees in the forest, understanding the combinatorial structure of such networks might therefore pave the way to extending current phylogeny-based HGT-inference methods in this direction. A central question in this context is, how can we construct an arboreal network? Answering this question is strongly informed by finding ways to \textit{encode} an arboreal network, that is, breaking up the network into simpler combinatorial structures that, in a well defined sense uniquely determine the network. In the form of triplets, trinets and quarnets such encodings are known for certain types of single-rooted phylogenetic networks. By studying the underlying tree of an arboreal network, we compliment them here with an answer for arboreal networks.

We posit that gene-environment interplay (GxE) studies should be developed both theoretically and empirically to be of relevance to policy makers. On the theoretical front, this development is essential because the current literature lacks the integration of a clear framework capturing the various goals of public policies. Empirically, GxE models need to be further developed because the common way of modelling GxE effects fails to adequately capture the heterogeneous effects public policies may have along the distribution of genetic propensities (as captured by polygenic indices). We fill these gaps by proposing a policy classification for GxE research and by offering guidance on advancing the empirical modelling of policy-informative GxE interplay. While doing so, we provide a systematic review of existing GxE studies on educational outcomes exploiting policy reforms or environments that could be targeted by public policy.

Forecasts of hospitalisations of infectious diseases play an important role for allocating healthcare resources during epidemics and pandemics. Large-scale analysis of model forecasts during the COVID-19 pandemic has shown that the model rank distribution with respect to accuracy is heterogeneous and that ensemble forecasts have the highest average accuracy. Building on that work we generated a maximally diverse synthetic dataset of 324 different hospitalisation time-series that correspond to different disease characteristics and public health responses. We evaluated forecasts from 14 component models and 6 different ensembles. Our results show that component model accuracy was heterogeneous and varied depending on the current rate of disease transmission. Going from 7 day to 14 day forecasts mechanistic models improved in relative accuracy compared to statistical models. A novel adaptive ensemble method outperforms all other ensembles, but is closely followed by a median ensemble. We also investigated the relationship between ensemble error and variability of component forecasts and show that the coefficient of variation is predictive of future error. Lastly, we validated the results on data from the COVID-19 pandemic in Sweden. Our findings have the potential to improve epidemic forecasting, in particular the ability to assign confidence to ensemble forecasts at the time of prediction based on component forecast variability.

The differentiation between pathological subtypes of non-small cell lung cancer (NSCLC) is an essential step in guiding treatment options and prognosis. However, current clinical practice relies on multi-step staining and labelling processes that are time-intensive and costly, requiring highly specialised expertise. In this study, we propose a label-free methodology that facilitates autofluorescence imaging of unstained NSCLC samples and deep learning (DL) techniques to distinguish between non-cancerous tissue, adenocarcinoma (AC), squamous cell carcinoma (SqCC), and other subtypes (OS). We conducted DL-based classification and generated virtual immunohistochemical (IHC) stains, including thyroid transcription factor-1 (TTF-1) for AC and p40 for SqCC, and evaluated these methods using two types of autofluorescence imaging: intensity imaging and lifetime imaging. The results demonstrate the exceptional ability of this approach for NSCLC subtype differentiation, achieving an area under the curve above 0.981 and 0.996 for binary- and multi-class classification. Furthermore, this approach produces clinical-grade virtual IHC staining which was blind-evaluated by three experienced thoracic pathologists. Our label-free NSCLC subtyping approach enables rapid and accurate diagnosis without conventional tissue processing and staining. Both strategies can significantly accelerate diagnostic workflows and support efficient lung cancer diagnosis, without compromising clinical decision-making.

Neurons encode information in a binary manner and process complex signals. However, predicting or generating diverse neural activity patterns remains challenging. In vitro and in vivo studies provide distinct advantages, yet no robust computational framework seamlessly integrates both data types. We address this by applying the Transformer model, widely used in large-scale language models, to neural data. To handle binary data, we introduced Dice loss, enabling accurate cross-domain neural activity generation. Structural analysis revealed how Dice loss enhances learning and identified key brain regions facilitating high-precision data generation. Our findings support the 3Rs principle in animal research, particularly Replacement, and establish a mathematical framework bridging animal experiments and human clinical studies. This work advances data-driven neuroscience and neural activity modeling, paving the way for more ethical and effective experimental methodologies.

Mathematical and computational modelling in oncology has played an increasingly important role in not only understanding the impact of various approaches to treatment on tumour growth, but in optimizing dosing regimens and aiding the development of treatment strategies. However, as with all modelling, only an approximation is made in the description of the biological and physical system. Here we show that tissue-scale spatial structure can have a profound impact on the resilience of tumours to immunotherapy using a classical model incorporating IL-2 compounds and effector cells as treatment parameters. Using linear stability analysis, numerical continuation, and direct simulations, we show that diffusing cancer cell populations can undergo pattern-forming (Turing) instabilities, leading to spatially-structured states that persist far into treatment regimes where the corresponding spatially homogeneous systems would uniformly predict a cancer-free state. These spatially-patterned states persist in a wide range of parameters, as well as under time-dependent treatment regimes. Incorporating treatment via domain boundaries can increase this resistance to treatment in the interior of the domain, further highlighting the importance of spatial modelling when designing treatment protocols informed by mathematical models. Counter-intuitively, this mechanism shows that increased effector cell mobility can increase the resilience of tumours to treatment. We conclude by discussing practical and theoretical considerations for understanding this kind of spatial resilience in other models of cancer treatment, in particular those incorporating more realistic spatial transport.

Prior results for tRNA and 5S rRNA demonstrated that secondary structure prediction accuracy can be significantly improved by modifying the parameters in the multibranch loop entropic penalty function. However, for reasons not well understood at the time, the scale of improvement possible across both families was well below the level for each family when considered separately. We resolve this dichotomy here by showing that each family has a characteristic target region geometry, which is distinct from the other and significantly different from their own dinucleotide shuffles. This required a much more efficient approach to computing the necessary information from the branching parameter space, and a new theoretical characterization of the region geometries. The insights gained point strongly to considering multiple possible secondary structures generated by varying the multiloop parameters. We provide proof-of-principle results that this significantly improves prediction accuracy across all 8 additional families in the Archive II benchmarking dataset.

In this paper, we present a simple method to integrate risk-contact data, obtained via digital contact monitoring (DCM) apps, in conventional compartmental transmission models. During the recent COVID-19 pandemic, many such data have been collected for the first time via newly developed DCM apps. However, it is unclear what the added value of these data is, unlike that of traditionally collected data via, e.g., surveys during non-epidemic times. The core idea behind our method is to express the number of infectious individuals as a function of the proportion of contacts that were with infected individuals and use this number as a starting point to initialize the remaining compartments of the model. As an important consequence, using our method, we can estimate key indicators such as the effective reproduction number using only two types of daily aggregated contact information, namely the average number of contacts and the average number of those contacts that were with an infected individual. We apply our method to the recent COVID-19 epidemic in the Netherlands, using self-reported data from the health surveillance app COVID RADAR and proximity-based data from the contact tracing app CoronaMelder. For both data sources, our corresponding estimates of the effective reproduction number agree both in time and magnitude with estimates based on other more detailed data sources such as daily numbers of cases and hospitalizations. This suggests that the use of DCM data in transmission models, regardless of the precise data type and for example via our method, offers a promising alternative for estimating the state of an epidemic, especially when more detailed data are not available.